The present invention relates to a computer information processing method, and more specifically, to a method and system for protecting specific information.
In the current database system or other big data systems, in addition to sensitive data of common digital, text formatting, user sensitive image information is increasingly stored in the form of an object such as a picture, and in particular, there is extremely sensitive personal information in an image such as a copy or a scanned copy of an ID card, a bank card, a medical record, etc. When such information is displayed in a bank, healthy and other industry data retrieval system, on one hand, identity and right of the owner of the image needs to be audited by means of content carried by the image so that they could be used in related business, e.g., ID authenticity of a user at a counter is determined by looking over a user's photo on the ID card and checking birth date on the ID card; on the other hand, the sensitive information is very likely to be inappropriately used.
Thus, there is a need for a method in which when most valuable and sensitive information (e.g., ID-type information such as identity card number, bank card number) is being transmitted, read, presented by a front-end system via an application program or Http, it will be automatically covered. Similar situation also exists during transmission via FTP or other file management system or internet, i.e., there is a risk of being attacked and abused when sensitive, personalized and private content is being transmitted, stored and presented. All current solutions are to learn content pattern of sensitive information to be protected, analyze the picture content, find out content that matches with a set pattern of content to be protected, and then cover the content. It is very difficult to search for precisely matching content in a picture, particularly for such image content with unfixed content as ID card number, credit card number, and the like; and it becomes more difficult to identify detailed semantic content in an image when there is a certain defect in the image itself.